其次，分析了数据挖掘中所使用的关联规则和序列模式，对关联规则和序列模式的各种挖掘算法进行了比较。
Secondly, it analyzed association rule and sequence mode used in the process of data mining and compared the main algorithms of association rule and sequence mode.
首先在具有明显边缘特征的图像序列中，提取图像强边缘块的直方图，然后，利用时域多分辨算法，比较不同帧之间的图像差异。
Firstly, the histogram is computed from the strong edge blocks in the image sequences. Then, different images are compared by algorithm of multi-resolution steps.
由于检测对DC序列进行，算法计算时间复杂度比较低。
Since the detection is carried out on DC sequence, the computation complexity of this method is relatively low.
实验结果表明：该算法能够较好地从视频序列中分割运动前景和背景，比较适合于在基于内容的视频编码标准MPEG - 4中使用。
Experiment results show that the algorithm can preferably segment moving foreground and background in video sequence and it fits for MPEG-4coding standard, which is content-based.
水印对剪切、JPEG有损压缩、中值滤波、抵抗噪声干扰等常规图像处理手段比较其他的随机序列扩频算法具有更好的鲁棒性。
Compared with the other spread spectrum algorithm, this one is more robust against the typical image processing attacks such as adding noise, filtering and JPEG compression.
对广义自缩序列生成器，利用猜测攻击的思想给出了一种比较快速的初态重构算法。
An initial reconstruction algorithm is given for the generalized self-shrinking sequences using the ideas of the guessing attack.
本文在研究当前比较流行的一些序列模式挖掘算法的基础上，重点分析了MEMISP算法的不足。
Based on some sequential pattern mining algorithm, the dissertation analyzed the shortage of the MEMISP algorithm, and proposed an improved MEMISP algorithm.
采用动态时间归整(DTW)算法对语音信号进行特征参数序列比较并识别出结果。
The speech recognition system adopts dynamic time warping (DTW) algorism to compare the characteristic parameters of speech signal with each other and recognize the result of speech signal.
在第一章，介绍了序列比较常用的两种方法，一是序列比对算法，二是基于矩阵不变量方法。
In the first chapter, we introduce two methods of sequences comparison, one is the alignment, and the other is the method of the matrix invariant.
提出了一种基于估价函数的启发式生成uio序列的算法，通过分析比较该算法能更有效地产生UIO序列。
A heuristic algorithm based on evaluation function for UIO sequence generation is proposed, this algorithm can efficiently generate UIO sequence.
本文对线性选择算法在比较算法类中给出了基于中值序列而设计的任何改进算法的复杂度下界；
The lower bound of worst complexity is presented for any improved BFPET selecting algorithm based on median sequence.
在比较网络模型的基础上，该算法使用二分法思想，利用双调序列，构造出了一种并行的排序算法：双调排序网络。
Base on comparing net model, it constructs a kind of pallel sorting algorithm bitonic sorting net, taking advantage of dichotomy ideas and bitonic serial.
最后，使用实际图像序列实验得出算法的平均误判率为3.4%，比较讨论和改进思路一并给出。
Finally, the experiments in several images show the error rate of ITC-Seg is 3.4%, discussions and future improvements of the method are also given.
通过和现有算法的比较以及在大量真实序列图像上的实验表明，该算法不但能够更准确地描述目标的大小，而且显著提高了跟踪算法的精度。
In the presented tracking examples, the new method can be used to describe the target more accurately and thus achieves much better tracking precision.
通过和现有算法的比较以及在大量真实序列图像上的实验表明，该算法不但能够更准确地描述目标的大小，而且显著提高了跟踪算法的精度。
In the presented tracking examples, the new method can be used to describe the target more accurately and thus achieves much better tracking precision.
应用推荐